***Subset models: Men vs. Women***

***Open Gendered Justice_main data in STATA 18**

***Table D1- Men only model***

logit libdecis_crim laglibpct i.const_crim i.proc_crim i.violent i.property i.drugs i.morality SA UK India Philip Canada if femdef==0, cluster(year)

fitstat

***Table D2 - Predicted probs, discrete change & percent change for men only***

margins const_crim, post  
lincom 1.const_crim - 0.const_crim, level(95)
display ((_b[1.const_crim] - _b[0.const_crim])/_b[0.const_crim])*100

quietly logit libdecis_crim laglibpct i.const_crim i.proc_crim i.violent i.property i.drugs i.morality SA UK India Philip Canada if femdef==0, cluster(year)
margins proc_crim, post  
lincom 1.proc_crim - 0.proc_crim, level(95)
display ((_b[1.proc_crim] - _b[0.proc_crim])/_b[0.proc_crim])*100

quietly logit libdecis_crim laglibpct i.const_crim i.proc_crim i.violent i.property i.drugs i.morality SA UK India Philip Canada if femdef==0, cluster(year)
margins violent, post  
lincom 1.violent - 0.violent, level(95)
display ((_b[1.violent] - _b[0.violent])/_b[0.violent])*100

quietly logit libdecis_crim laglibpct i.const_crim i.proc_crim i.violent i.property i.drugs i.morality SA UK India Philip Canada if femdef==0, cluster(year)
margins property, post  
lincom 1.property - 0.property, level(95)
display ((_b[1.property] - _b[0.property])/_b[0.property])*100

quietly logit libdecis_crim laglibpct i.const_crim i.proc_crim i.violent i.property i.drugs i.morality SA UK India Philip Canada if femdef==0, cluster(year)
margins drugs, post  
lincom 1.drugs - 0.drugs, level(95)
display ((_b[1.drugs] - _b[0.drugs])/_b[0.drugs])*100

quietly logit libdecis_crim laglibpct i.const_crim i.proc_crim i.violent i.property i.drugs i.morality SA UK India Philip Canada if femdef==0, cluster(year)
margins morality, post  
lincom 1.morality - 0.morality, level(95)
display ((_b[1.morality] - _b[0.morality])/_b[0.morality])*100


***Table D3 - Women only model***

logit libdecis_crim laglibpct i.const_crim i.proc_crim i.violent i.property i.drugs i.morality SA UK India Philip Canada if femdef==1, cluster(year)

fitstat

***Table D4 - Predicted probs, discrete change & percent change for women only***

margins const_crim, post  
lincom 1.const_crim - 0.const_crim, level(95)
display ((_b[1.const_crim] - _b[0.const_crim])/_b[0.const_crim])*100

quietly logit libdecis_crim laglibpct i.const_crim i.proc_crim i.violent i.property i.drugs i.morality SA UK India Philip Canada if femdef==1, cluster(year)
margins proc_crim, post  
lincom 1.proc_crim - 0.proc_crim, level(95)
display ((_b[1.proc_crim] - _b[0.proc_crim])/_b[0.proc_crim])*100

quietly logit libdecis_crim laglibpct i.const_crim i.proc_crim i.violent i.property i.drugs i.morality SA UK India Philip Canada if femdef==1, cluster(year)
margins violent, post  
lincom 1.violent - 0.violent, level(95)
display ((_b[1.violent] - _b[0.violent])/_b[0.violent])*100

quietly logit libdecis_crim laglibpct i.const_crim i.proc_crim i.violent i.property i.drugs i.morality SA UK India Philip Canada if femdef==1, cluster(year)
margins property, post  
lincom 1.property - 0.property, level(95)
display ((_b[1.property] - _b[0.property])/_b[0.property])*100

quietly logit libdecis_crim laglibpct i.const_crim i.proc_crim i.violent i.property i.drugs i.morality SA UK India Philip Canada if femdef==1, cluster(year)
margins drugs, post  
lincom 1.drugs - 0.drugs, level(95)
display ((_b[1.drugs] - _b[0.drugs])/_b[0.drugs])*100

quietly logit libdecis_crim laglibpct i.const_crim i.proc_crim i.violent i.property i.drugs i.morality SA UK India Philip Canada if femdef==1, cluster(year)
margins morality, post  
lincom 1.morality - 0.morality, level(95)
display ((_b[1.morality] - _b[0.morality])/_b[0.morality])*100